Live, Dynamic Salesforce Search

Overview

Our AI Agents platform is integrated seamlessly with Salesforce, unlocking a powerful capability: the ability to run real-time, dynamic SOQL (Salesforce Object Query Language) queries based on user prompts. Unlike conventional search tools, this integration goes beyond simple keyword matching or predefined filters - it intelligently understands user intent, dynamically constructs SOQL queries, and executes them in real time.

This enables your team to interact with Salesforce using natural language while still leveraging the full power of structured database querying.

Key Features

1. Dynamic SOQL Generation from Natural Language

Users can ask questions in plain English (or any other language), and our agents will generate and execute precise SOQL queries that match the request. This includes support for:

  • Custom fields and objects: All Salesforce customizations are respected.

  • Object relationships: Queries automatically incorporate JOIN-like logic via relationship fields.

  • Filters, sorting, and grouping: Use of WHERE, ORDER BY, and aggregate functions is supported when relevant.

Example prompts:

  • “Which accounts have multiple active support cases right now?”

  • “Find all opportunities that are not closed-won or closed-lost and rank them by ARR.”

  • “List all open support cases with high priority.”

The system interprets these requests in real time, dynamically generating SOQLs like:

SELECT Id, Name, CloseDate, TCV__c, Account.Name
FROM Opportunity
WHERE (CloseDate >= 2025-10-01) AND (CloseDate <= 2025-12-31)
ORDER BY TCV__c DESC NULLS LAST LIMIT 1

SELECT Id, CaseNumber, CreatedDate, Owner.Name
FROM Case
WHERE (IsClosed = false) AND (Account.Name = AccountID)
ORDER BY LastModifiedDate DESC NULLS LAST LIMIT 15

2. Real-Time Query Execution

SOQL queries are executed live against your Salesforce data, reflecting the most current state of records. There is no dependency on stale data caches or outdated index layers. This is particularly useful for:

  • Operational dashboards

  • Sales enablement tools

  • Customer success platforms

  • Executive reporting agents

3. Full Schema Awareness

Unlike general-purpose RAG (retrieval-augmented generation) systems that lack structural context, our agents:

  • Understand your full Salesforce schema, including custom objects and fields.

  • Recognize how entities are related via lookup, master-detail, and junction relationships.

  • Adapt automatically if your Salesforce instance is customized over time.

4. Cross-Object Reasoning

Our AI agents can answer multi-object questions that require reasoning across your Salesforce model. For example:

  • “Which opportunities over $50k are currently in MQL or later stages?”

  • “Which customers have open cases AND are in a renewal opportunity stage?”

These questions are translated into multi-object SOQL joins (e.g., Opportunity, Account, Case, Lead) and executed as a single query.

Example Use Cases

Prompt
What Happens Behind the Scenes

"List open support cases for Acme Corp"

Finds the AccountId for "Acme Corp", constructs a SELECT on Case with Status = 'Open' and Account.Name = 'Acme Corp'

"Show opportunities closing this quarter"

Calculates current quarter dates, filters CloseDate, and filters opportunities accordingly

"Who are our top ARR customers with active tickets?"

Aggregates Opportunity.Amount by Account, filters Case.Status = 'Open' and sorts results

Summary

Our Salesforce integration enables AI Agents to:

  • Understand your schema and relationships in Salesforce.

  • Translate natural language into accurate SOQL in real time.

  • Execute those queries securely and reflect up-to-date information.

  • Respect your customizations and user permissions.

This empowers your team to extract insights, triage issues, and make data-driven decisions - all by simply asking.

PreviousDots - Knowledge NuggetsNextDual-Layer Permissions Enforcement

Last updated

Was this helpful?